Project Summary/Abstract: Cancer genome sequencing studies have now revealed that genes that encode nine different subunits of SWI/SNF chromatin remodeling complexes are frequently mutated in a wide variety of human cancers. These include cancers of brain, ovary, breast, kidney, lung, pancreas, uterus, bladder, stomach, colon, liver, skin and blood. Collectively, over twenty percent of all human cancers contain a SWI/SNF mutation making SWI/SNF complexes the most frequently mutated chromatin/epigenetic regulator in cancer. During the current funding cycle we have made notable progress in elucidating the functions of SWI/SNF complexes. However, major questions have subsequently emerged. SWI/SNF complexes consist of both core subunits and variant subunits, with the latter present in only sub-classes of complexes. It has now become clear that the frequently mutated subunits are all variant subunits including ARID1A, SMARCA4 and PBRM1 and that mutation of each is associated with a distinct cancer spectrum. However, the mechanistic and functional contributions of these variant subunits and sub-classes to SWI/SNF function is poorly understood. We hypothesize that oncogenesis occurs not due to broad loss of SWI/SNF complex function but rather due to aberrant function of residual SWI/SNF complexes. We further hypothesize that loss of the variant tumor-suppressor subunits alters the composition, targeting and chromatin remodeling activity of SWI/SNF thus impairing differentiation and promoting oncogenesis. Using our genetically engineered primary cells, cell lines and mice, we will address three aims: Aim 1: How do the mutually exclusive ATPase subunits of SWI/SNF complexes, SMARCA4/BRG1 and SMARCA2/BRM, differ in function and what is the mechanistic basis for the synthetic lethality of SMARCA2 in SMARCA4 mutant cancers? Aim 2: How does the PBRM1-containig PBAF sub-class of SWI/SNF complexes differ from the ARID1A/B-containing BAF sub-class with respect to composition, targeting, chromatin remodeling activity, enhancer regulation and control of lineage specification? Aim 3: How does our newly discovered BRD9- containing sub-class differ in composition and function from other SWI/SNF sub-classes, and can BRD9 be exploited as a therapeutic target?